There have been only a few epidemiologic studies of risk factors for cancer in black Americans. It has been commonly assumed that information from studies in whites can be generalized to blacks, but this may not be the case. Furthermore, the effectiveness of health promotion and education efforts among blacks may be hampered by the lack of specific evidence about modifiable risk factors from studies of cancer in blacks. In addition, this will strengthen belief in causality. If they are not confirmed, this may lead to identification of the true causal factors. We propose to carry out a series of case-control studies of breast cancer, lung cancer, and large bowel cancer among black women and men. The proposed studies will make use of an already existing large database which has the relevant information on exposures, covariates, and outcomes; it provides the opportunity to learn about risk factors for cancer in blacks at minimal cost. Approximately 800 cases of breast cancer, 275 cases of lung cancer, 300 cases of large bowel cancer, and several thousand controls with other diagnoses will be studied. The incidence of breast cancer is rising faster among black women than white women. We will examine the relation of breast cancer risk to alcohol consumption, oral contraceptive use, induced abortion, cigarette smoking, body size, and lactation, all factors that are influenced by the economic, social, and cultural environment, and that may differ between blacks and whites. The results may help to clarify the etiology of breast cancer and may be useful for cancer prevention in black women. While rates of smoking have declined among blacks, the decline has not been as great as among whites. Recent cigarette marketing campaigns have been specifically aimed at the black community. We will assess cigarette smoking and lung cancer among blacks, including an exploration of the effect of intensity and duration of smoking and of differences in the type of cigarette smoked (tar content and filter). The results of the proposed study may have a greater impact than earlier general reports on the decisions of black Americans regarding smoking. We will also assess two hypotheses regarding large bowel cancer, one of which concerns an adverse effect of alcohol consumption and the other a protective effect of nonsteroidal anti- inflammatory drug use. Both exposures are common and modifiable.